Using panel data from 92 rural districts over 2000-2018, this paper examines the transmission channels through which government
spending affects agricultural growth and poverty in rural Uganda. A structural equation model that combines district-level public
expenditure records, crop‐specific agricultural output, household consumption surveys and sub-national poverty maps is estimated
in a system-GMM framework to address endogeneity and spatial spill-overs. Results show that a 1 % increase in public agricultural
expenditure raises agricultural GDP per rural worker by 0.34 % after two seasons, mainly through higher fertiliser use and extension
contacts. The elasticity of the rural poverty headcount to agricultural growth is –0.82, indicating that roughly 490 000 Ugandans are
lifted above the national poverty line for every percentage point of agricultural growth. Decomposition exercises reveal that 63 % of
the poverty reduction is mediated by higher real agricultural wages and 25 % by lower food prices; the remainder operates via
increased non-farm rural employment stimulated by agricultural demand linkages. Marginal benefit–cost ratios exceed 3.7 for
extension, 2.9 for feeder-road maintenance and 2.2 for rural education, while input subsidies exhibit ratios below unity once crowd-
out of private purchases is accounted for. Simulation of a 10 % reallocation of spending toward extension and rural roads suggests a
potential additional 1.8 percentage-point reduction in the rural poverty rate by 2025 without increasing the overall budget envelope.
The findings underscore the centrality of agricultural growth in Uganda’s rural poverty reduction strategy and provide evidence-based
guidance for prioritising public investments under fiscal constraints.